proposed moral motivational orientations. In order to do so, Study 1 examined the predicted relations between a measure based largely on existing regulatory focus scales but applied specifically to the moral domain and other established constructs in the literature (e.g., general regulatory focus, moral identity, and self-motives; H2). I
predicted that both an ought and an ideal orientation would positively correlate with the general measure of MID (the measure typically employed in the literature). However, I anticipated that only an ought orientation would positively correlate with a general prevention focus, self-maintenance motives, and having a stricter standard for engaging in the prosocial behaviors measured. In contrast, I predicted that only an ideal orientation would positively correlate with a general promotion focus, self-improvement motives, and a lower perceived standard for engaging in prosocial behaviors. I also predicted that engaging in prosocial behaviors would be perceived to be more intrinsically motivated for those with a stronger ideal (vs. ought) orientation. 4 Study 1 also looked at how the
moral motivational orientations related to prosocial behavior and to feelings and
emotions regarding the success and failure of moral actions (H3 and H5). Consistent with
the past literature, I predicted that the general measure of MID would positively correlate with self-reported amounts of prosocial behaviors (i.e., the stronger one’s MID, the more time volunteered). However, I predicted that ideal and ought motivational orientations would differ, such that ideal orientation would be a stronger, positive predictor of self- reported amounts of prosocial behavior (H4). I also expected to see differences between
an ought vs. ideal orientation with respect to positive and negative feelings toward past
4 An ought-oriented individual is motivated to do the right thing and avoid doing the wrong thing because it
is perceived as a moral obligation. Thus, when an individual does the right thing or refrains from doing the wrong thing, it is because they feel compelled to abide by the moral law (an extrinsic source). The key issue is the lack of autonomy (perceived choice) over following one’s moral beliefs. A feeling of autonomy is a critical component of self-determined, and thus, intrinsically motivated behavior. Since maintaining one’s moral integrity concerns perceived duties, there is no room for personal choice and preference.
In contrast, an ideal-oriented individual is motivated to do moral good and avoid doing immoral things, because they perceive it as improving their moral self (i.e., going above and beyond their moral self- expectations). They are doing things that they don’t normally do, or don’t feel obligated to do. Thus, personal choice is involved, as they must actively go above their moral standards to become a better person. Janoff-Bulman et al. (2009) also supports this distinction, as they found that people considered prescriptive moral behaviors (moral ideals) to be more optional and open to one’s personal preference, whereas proscriptive moral behaviors (moral oughts) were rated as more mandatory, or less discretionary.
and future prosocial behaviors. In particular, I predicted that an ought orientation would more strongly predict active (vs. passive), negative feelings in reaction to failures to engage in prosocial behaviors. In contrast, I predict that an ideal orientation would more strongly predict energizing and active (vs. passive), positive feelings toward successes in engaging in prosocial behaviors.
Methods
Participants & procedure. Ninety-eight participants were recruited through Amazon Mechanical Turk and filled out the survey via Qualtrics. Participants first completed a series of measures that assessed their general MID (i.e., what has typically been measured in the literature), the hypothesized moral orientations, and measures that moral ought and ideal orientations were expected to correlate with (e.g., prevention vs. promotion focus, self-improvement vs. self-maintenance motives, and proscriptive vs. prescriptive self-standards toward engaging in prosocial behaviors). Following this, participants answered questions concerning their engagement in past prosocial behaviors (e.g., time spent volunteering per month), and also their feelings toward their success and failure to engage in prosocial behaviors. While there are a large variety of prosocial behaviors to potentially measure, Reed, Aquino, & Levy (2007) suggested that
volunteering time resonates more with high moral identifiers than donating the equivalent amount of money or resources because it is seen as more self-expressive and effortful.
Measures.
General moral identity. The go-to measure for MID is Aquino & Reed’s (2002) 10-item scale of moral identity, and as such, their scale was used to measure an
between internalized vs. symbolized moral identity, which differentiates between more personal/private vs. social/public identity concerns. Each subscale was assessed through 5 items. For example, the internalized (private) subscale contains items such as, “It would make me feel good to be a person who has these characteristics” and “I strongly desire to have these characteristics” (α = .67), whereas the symbolized (public) subscale contains items such as, “I often wear clothes that identify me as having these
characteristics” and “I am actively involved in activities that communicate to others that I have these characteristics” (α = .85). Each item was scored on a scale from 1 (strongly disagree) to 5 (strongly agree) (see Appendix A for the full scale).
Ought vs. ideal moral orientations. To develop a measure of ought vs. ideal moral orientations I drew on existing promotion vs. prevention focus measures (e.g., Lockwood, Jordan, & Kunda, 2002; adapted to the moral domain, of course) and
conceptions of ought vs. ideal selves (e.g., Higgins, Klein, Strauman, 1985). I created a new 10-item measure that was intended to capture the relative strength of one’s ought vs. ideal moral orientation. Half of the items were designed to capture an ought orientation (α = .82), focusing on maintaining one’s moral self by avoiding becoming a worse person (e.g., “Doing the right thing is important to who I am, primarily because it keeps me from
becoming a worse person”, “In order to remain a good person, I feel like I must help others”, “As a central part of who I am, I have a responsibility to do good deeds”). The
other half captured an ideal orientation (α = .86), focusing on improving one’s moral self by becoming a better person (e.g., “Doing the right thing is important to who I am,
primarily because it makes me a better person”, “In order to become a better person, I strive to help others”, “As a central part of who I am, I have aspirations to do good
deeds”). See Appendix B for the full scale.5 Participants rated the degree to which each item represented them on a 9-point scale (1 = not me at all, 9 = completely me).
Regulatory focus. General promotion vs. prevention focus was measured using Lockwood’s et al. (2002) scale adapted for general life goals. The scale was comprised of 18 items, 9 that tapped a promotion focus (e.g., “I frequently imagine how I will achieve my hopes and aspirations”, α = .89), and 9 that tapped a prevention focus (e.g., “I am anxious that I will fall short of my responsibilities and obligations”, α = .91). Each item was rated on a 9-point scale, ranging from 1 (not at all true of me) to 9 (very true of me). See Appendix C for the full scale.
Self-motives. Self-improvement and self-maintenance motives were measured by 12 items (adapted from Breines & Chen, 2012). Six items tapped self-improvement (e.g., “I am primarily concerned about improving who I am”; “Making progress, as a person, is the most important thing to me”, α = .91), and 6 items tapped self-maintenance (e.g., “I am primarily concerned about maintaining who I am”; “Being consistent with my sense of self is what I desire the most”, α = .79). Each question was rated on a 9-point scale, ranging from 1 (not at all true of me), to 9 (very true of me). See Appendix D for the full scale.
Prosocial self-standard. The extent to which participants held themselves to a proscriptive (strict) or prescriptive (forgiving) regarding engagement in prosocial
5 After rating all of the items individually, the ought and ideal items were broken into pairs, and
participants were asked to pick which item in each pair best represented them. Thus, the measure provided two values: the relative strength of each orientation based on scale ratings, and the proportion of ought vs. ideal statements chosen in a dichotomous fashion. There are multiple ways that an
individual’s ought vs. ideal orientation can be calculated with these values. Given that the dichotomous choices provided much less nuanced information regarding ought and ideal orientations, I decided to go
behaviors was measured with 3 items that asked about participants’ expectations and self- standards toward volunteerism (e.g., “I have a strong expectation for myself to engage in this behavior”; “I feel like I have a duty to engage in this behavior”; “Simply put, I expect myself to engage in this behavior”, α = .89). Each item was rated on a 9-point scale, ranging from 1 (strongly disagree) to 9 (strongly agree).
Intrinsic vs. extrinsic motivation. The extent to which engaging in prosocial behaviors was intrinsically vs. extrinsically motivated was measured with an adapted version of Mullan, Markland, & Ingledew’s (1997) Behavioral Regulation in Exercise Questionnaire (BREQ) scale. The overall measure was comprised of 2 subscales:
external regulation (e.g., “I donate because other people say I should”, “When I
volunteer/donate, it's often because I feel guilt-tripped (by others, or myself) into doing it”, α = .84) and intrinsic regulation (“I find donating a pleasurable experience”, “I get a lot of enjoyment and satisfaction from volunteering/donating”, α = .93). The 9 items were rated on a 5-point scale, ranging from 0 (not true for me) to 4 (very true for me). See Appendix E.
Prosocial behavior. Prosocial behavior was measured by asking participants to estimate the average amount of time (in hours) that they spend per month volunteering at non-profit organizations, charities, or events.
Feelings toward prosocial behavior. Participants’ feelings toward past prosocial behaviors (both successes and failures) were assessed with an 18-item measure, with 9 positive feelings (e.g., Praiseworthy, Satisfied, Pleasure, etc.), and 9 negative feelings (Blameworthy, Displeased, Guilty, etc.). For positive feelings, participants were asked “Thinking about the times that you have already volunteered for or donated to charities
and non-profit organizations, to what extent did it/does it make you feel the following?”. For the negative items, participants were asked, “Thinking about the times that
you decided not to volunteer for or donate to charities and non-profit organizations, to what extent did it/does it make you feel the following?”. All items were rated on a 9- point scale ranging from 1 (not at all) to 9 (completely), see Appendix F for the full scale. The inclusion of more energetic emotions (e.g., Alive and vitalized; Full of energy and spirit; Energized – adapted from Ryan & Frederick’s, 1997 Subjective Vitality Scale) allowed for an initial test of the idea that those with an ideal (vs. ought) moral orientation would be more energized by successful moral behaviors.
Results
Preliminary analyses. A Principal Components Analysis (PCA) with an oblique rotation (Direct Oblimin) was used to assess the 12-item moral orientation measure. While the analysis did pull out two components (based on Eigenvalues > 1 and the break point in the scree plot of eigenvalues) accounting for 61.8% of the variance, the
components did not neatly divide into the intended ought and ideal components. The pattern and structure matrices showed that two of the intended ought items (“As a central part of who I am, I have a responsibility to do good deeds”, “Doing the right thing is mostly driven by my desire to meet my standards for myself”) loaded > .60 onto an ideal- like component, and two of the intended ideal items (“Doing the right thing is important to who I am, primarily because it makes me a better person”, “In order to become a better person, I strive to help others”) loaded > .70 onto an ought-like component. It is not entirely obvious why these items loaded onto the ‘wrong’ components, but one
have been too easy for participants to gloss over the subtle differences and to treat two items as the same. For example, one of the ought items was “As a central part of who I am, I have a responsibility to do good deeds”, whereas the ideal item equivalent was “As a central part of who I am, I have aspirations to do good deeds” (italics added). It seems plausible that participants focused on the fact that both were talking about doing good deeds, and did not pay much attention to the words ‘responsibility’ and ‘aspirations’. In an attempt to remedy this issue, I dropped the four items that loaded > .50 onto the opposite components, and reran the PCA. Again, the analysis revealed two components that accounted for 65% of the variance. This time, the pattern and structure matrices showed that the four remaining ought items loaded > .69 onto one component, and the four remaining ideal items loaded > .63 onto the other component.6 The Cronbach’s
Alpha for the reduced set of ought and ideal items were .82 and .79 (respectively), indicating good reliability. Ought and ideal orientation scores were modestly correlated with each other (r = .54).
Ought vs. ideal orientations. The following analyses are based on linear regressions in which ought and ideal motivational orientation scores were entered simultaneously.
MID. As predicted, both ought (β = 0.27, t(95) = 2.56, p = .012) and ideal (β = 0.31, t(95) = 2.93, p = .004) orientations significantly and positively predicted the general
6 I ran all the analyses using both the original ought and ideal scales (i.e., using all of the items), and the
shorter version. Despite the messy factor analysis when all items were included, the analyses looked very similar whether the original scales were used, or the short version were used. The only differences were that ought orientation marginally predicted general MID (p = .08) using the original scales (vs. p = .012 using the short version), and only ideal (but not ought) orientation predicted positive feelings (p < .001 and p = .26) using the original scales, whereas both ideal and ought orientations predicted positive feelings with the short version (ps < .004).
measure of MID. Interestingly, though, when MID was broken down into the
symbolization and internalization subcomponents (Aquino & Reed, 2002), the two types of moral orientation differed in which subcomponent of MID they significantly predicted. Specifically, only ought orientation strongly predicted symbolization, (β = 0.39, t(95) = 3.76, p < .001), whereas ideal orientation was only marginally related, β = 0.20, t(95) = 1.92, p = .057. Likewise, only ideal orientation significantly predicted internalization (β = 0.29, t(95) = 2.48, p = .015), whereas ought orientation did not, β = -0.12, t(95) = -1.06,
p = .29. Although I had no a priori predictions concerning these subcomponents of MID,
the results are congruent with the hypothesized distinction between ought and ideal orientations. Specifically, the symbolization subcomponent of MID captures more of an extrinsically driven form of moral identity (e.g., a more publically visible MID), which is consistent with my conception of an ought orientation. In contrast, the internalization subcomponent of MID captures more of a private and internally focused MID, which is consistent with my conception of an ideal orientation.
Self-motives. In support of my predictions, ought orientation significantly predicted self-maintenance scores (β = 0.48, t(95) = 4.61, p < .001), but ideal orientation did not, β = 0.06, t(95) = 0.58, p = .56. In contrast, ideal orientation predicted self- improvement scores (β = 0.49, t(95) = 4.93, p < .001), but ought orientation did not, β = 0.12, t(95) = 1.21, p = .23. This pattern suggests that, as hypothesized, an ought
motivational orientation is associated with consistency type motives, whereas an ideal orientation is associated with change and improvement type motives.
Regulatory focus. Consistent with my predictions, ought orientation significantly and positively predicted general prevention focus (β = 0.63, t(95) = 6.13, p < .001), while
ideal orientation negatively predicted prevention focus, β = -0.39, t(95) = -3.78, p < .001. For promotion focus, only ideal orientation was a significant predictor (β = 0.61, t(95) = 6.45, p < .001), while ought orientation was not, β = 0.01, t(95) = 0.15, p = .86.
Perceived standard. The extent that participants held themselves to a strict personal standard for volunteering was independently predicted by both ought and ideal orientations, β = 0.27, t(95) = 2.54, p = .01 and β = 0.30, t(95) = 2.83, p = .006,
respectively. This was inconsistent with my predictions, since I anticipated that those with an ideal orientation would construe volunteering as more of a matter of preference compared to ought-oriented individuals.
Locus of motivation. Supporting my predictions, ought orientation was a significant predictor of extrinsic motivation to volunteer (β = 0.37, t(95) = 3.28, p = .001), while ideal orientation was not, β = -0.05, t(95) = -0.43, p = .67. For intrinsic motivation to volunteer, both ought and ideal orientations were significant predictors, however, the relative strengths of the associations were consistent with my predictions. Specifically, ideal orientation had a stronger association with intrinsic motivation (β = 0.58, t(95) = 6.52, p < .001, ∆R2 = .24) than ought orientation did, β = 0.18, t(95) = 2.05,
p = .04 (∆R2 = .02).
Prosocial feelings and behaviors.
Volunteerism. While I predicted that both ought and ideal orientations would significantly predict self-reported hours (per month) of volunteerism, only ideal
orientation was a significant predictor (β = 0.36, t(94) = 3.16, p = .002), whereas ought orientation was not, β = -0.01, t(94) = -0.10, p = .92. In a separate analysis, I also predicted that the general measure of MID would significantly predict volunteerism,
which it did, β = 0.30, t(95) = 3.05, p = .003. In a third analysis, I also examined
associations between moral orientation and the subcomponents of MID and volunteerism and found that only symbolization (but not internalization) was a significant predictor, β = 0.28, t(94) = 2.82, p = .006 and β = 0.09, t(94) = 0.91, p = .37 (respectively). This was interesting because while symbolization predicted volunteerism, ought orientation (which was associated with symbolization) did not. This suggests that while ought and ideal orientations were associated with the symbolization and internalization subcomponents of MID (respectively), they are not the same constructs. When accounting for all of the variables (ought orientation, ideal orientation, general MID, symbolization, and
internalization) in a single analysis, only ideal orientation remained a significant predictor of volunteerism, β = 0.31, t(91) = 2.56, p = .01. All other variables were non-significant,
ps > .52.
Feelings. While I originally predicted a difference between active and passive feelings (both positive and negative) regarding past success and failures of volunteerism for the different moral orientations, a PCA with Direct Oblimin rotation revealed that all the positive emotion items (regardless of whether they were active vs. passive) loaded onto a single factor, and all the negative items loaded onto a single factor (based on eigenvalues > 1 and the break point of the scree plot of eigenvalues). Thus, I combined all the positive items into a single positive feelings measure, and all the negative items into a single negative feelings measure. Both ought and ideal orientation significantly predicted positive feelings toward past successful volunteerism, β = 0.29, t(95) = 2.92, p = .004 and β = 0.37, t(95) = 3.74, p < .001 (respectively). Thus, participants seemed to experience equally positive feelings for past volunteerism, regardless of their moral
orientation. However, only ought orientation significantly predicted negative feelings toward past failures to volunteer (β = 0.36, t(95) = 3.14, p = .002), whereas ideal
orientation was not a significant predictor, β = -0.10, t(95) = -0.83, p = .41. This pattern of results is broadly consistent with my predictions because individuals with an ought orientation were hypothesized to be more concerned about failing their moral goals than those with an ideal motivational orientation.7
Discussion
The moral orientation measure designed for this initial study turned out to be a little messy. However, dropping four items resulted in the remaining eight items loading onto factors representing an ought vs. an ideal orientation. Overall, however, the shorter version of the measure (and, surprisingly, the original measure as well) provided
evidence consistent with my conceptualization of two moral motivational orientations.
7 To demonstrate that the moral orientations are not redundant with other established moral constructs, I
examined their relationship to identification with all of humanity (IDAH) and empathic concern. Ought and ideal orientation both significantly predicted IDAH scores, β = 0.31, t(95) = 2.91, p = .005 and β = 0.26, t(95) = 2.41, p = .018 (respectively). Given that the two moral orientations were distinctly associated with different outcomes, it suggests that IDAH is somewhat unique from ought and ideal orientations. I also ran the key analyses above while controlling for IDAH (i.e., the two moral orientations and IDAH were entered simultaneously). The results were the same, with ideal orientation uniquely predicting
volunteerism, self-improvement, promotion focus, positive emotions, and intrinsic motivation (ps < .02), and only ought orientation predicting self-maintenance, negative emotions, and extrinsic motivation (ps < .01). Similar to the previous analyses, both orientations predicted prevention focus (ps < .001), with a positive relationship for ought orientation (β = 0.65) and a negative relationship for ideal orientation (β = - 0.37).
In contrast to IDAH, only ideal orientation significantly predicted empathic concern (β = 0.32, t(95) = 2.79, p = .006), whereas ought orientation was non-significant, β = 0.07, t(95) = 0.65, p = .52. However, ideal orientation and empathic concern were only modestly correlated with each other (r = .36), suggesting that the two factors, while related, were not the same. I ran all the key analyses controlling for empathic concern (i.e., both moral orientations and empathic concern were entered simultaneously). Similar to the analyses controlling for IDAH, including empathic concern into the regression analyses did not affect the results. Ideal motivational orientation still uniquely predicted volunteerism, self-
improvement, promotion focus, and intrinsic motivation (ps < .006). In contrast, ought motivational orientation uniquely predicted self-maintenance, negative emotions, and extrinsic motivation (ps < .003). Both orientations predicted prevention focus (ps < .001), but the relationship was positive for ought orientation (β = 0.63) and negative for ideal orientation (β = -0.39).
Ought and ideal subcomponents were differentially and significantly associated with the majority of the other measures that were predicted to relate to the two moral orientations in different ways (H1). More specifically, an ought moral orientation was uniquely